1. Field Of The Invention
This invention relates to innovations in the design of bunching (also called prebunching) cavities for high power gyroklystron amplifiers or phase-locked gyroklystron oscillators, and more particularly to the arrangement of slots in the walls of the bunching cavities as well as in the cut-off regions.
2. Background Description
A gyroklystron is a cyclotron maser device operating in the gyrotron mode (i.e., either near cutoff in a microwave cavity, or with k.sub.z of the mode near zero) that employs one or more bunching cavities, separated by drift spaces that are cutoff to the modes of the bunching cavities, and followed by an output cavity. As such, it operates in a strong axial magnetic field, such that the operating frequency is near the cyclotron frequency or one of its harmonics. An external signal is applied to the first of the bunching cavities, and used to initiate a phase-modulation of the beam. This modulation is magnified by transit through the drift spaces (much as in a conventional klystron the axial velocity modulation leads to axial bunching). The output cavity acts either as an amplifier of the external signal, in which case the device is called a gyroklystron amplifier, or alternatively, the output cavity will produce power, i.e. oscillate, without an external signal, in which case one can attempt to phase lock and frequency lock this oscillation and the device is called a phase-locked gyroklystron oscillator.
There are a number of design constraints with respect to the bunching cavities of a gyroklystron amplifier or phase-locked gyroklystron oscillator. Specifically, there are two pairs of conflicting constraints that a design must satisfy. The first pair of conflicting constraints are that the cavity (or cavities) must be stable against self-oscillation both in the desired mode, e.g. the fundamental, cylindrical TE.sub.111 mode, as well as in competing axial and transverse modes, in the first and higher harmonics of the cyclotron maser interaction, while simultaneously sustaining large drive fields from an external source in order to produce the bunching of the electron beam that permits amplifier or phase-locked oscillator operation. The second pair of conflicting constraints are that the bunching cavity or cavities must be isolated from each other and from the output cavity, so that information will not flow back from the output cavity to the bunching cavities, causing the system to self-oscillate or oscillate without phase control, while at the same time the diameter of the drift spaces must permit transit of the electron beam that drives the interaction. This constraint becomes more difficult to achieve at higher frequencies, e.g. 35 GHz and above, due to the necessity that the transverse dimension of the drift space be below cutoff to the operating mode. For instance, if the operating mode is the fundamental rectangular mode, the transverse dimension of the drift space would be less than or on the order of 1/2 of the free-space wavelength of the mode. If the operating mode is the fundamental cylindrical mode, the transverse dimension of the drift space would be less than or on the order of .586 of the free-space wavelength of the mode. Furthermore, precise tuning of the bunching cavity with respect to the output cavity is essential, so that the ability to mechanically tune the cavity in order to obtain this precise tuning without remachining of the cavity is very valuable.
Previous phase-locked gyroklystron oscillator bunching cavities have faced the same design constraints, and the cavity designs employed had severe limitations. Cavity loading was accomplished by means of resistive walls, which are very inflexible in determining ultimate cavity Q-factors. Furthermore, the previous method did not provide control of competing transverse modes by preferentially lowering their Q-values. It also did not provide a means to control the length of the interaction both in the lowest order axial mode (the preferred mode) and in higher order axial modes, thus increasing the danger of self-oscillation in these modes, which would prevent successful gyroklystron operation. The disadvantages of the old approach apply particularly to devices designed to operate at millimeter-wave and higher frequencies, where the device cross sections decrease, making the twin requirements of beam propagation and cavity cutoff difficult to simultaneously satisfy.
The foregoing illustrates limitations known to exist in present devices. Thus, it is apparent that it would be advantageous to provide an alternative directed to overcoming one or more of the limitations set forth above. A suitable alternative is provided including features more fully disclosed hereinafter.